Biochcmical Pharmacology, Vol. 26, pp. 947 950. Pergamon Press. 1977. Printed in Great Britain.

RETINOID ACID ESTERASE ACTIVITIES: TISSUE A N D SUBCELLULAR DISTRIBUTION IN

MICE*

CHIH-CHENG WANG and DONALD L. HILL Kettering-Meyer Laboratory, Southern Research Institute, Birmingham, AL 35205, U.S.A.

(Received 30 June 1976: accepted 15 October 1976)

Abstract--Methyl retinoate, ethyl retinoate and a trimethylmethoxyphenyl (TMMP) analog of ethyl retinoate are cleaved by enzymes widely distributed in tissues of mice; brain, liver and ovary contain high activity and plasma contains low activity for all three substrates. In the lung, enzymatic activity is primarily localized in the microsomal fraction, with some activity also in the soluble cytoplasmic fraction. From studies on subcellular distribution, the effect of pH, sensitivity to inhibitor and thermal stability, it appears that there are several esterases that act upon retinoid acid esters.

Retinoids (vitamin A compounds and their synthetic analogs) are potent agents for controlling differentia- tion and growth and for preventing the process of chemical carcinogenesis in several different epithelial tissues including lung [1-3]. However, the usefulness of natural retinoids as anticarcinogenic agents is limited by excessive toxicity [3]. Since the undesirable toxic effects may be associated with the presence of a free terminal carboxyl group, carboxyl esters of retinoic acid and an ester of a aromatic retinoic acid analog have been synthesized in an attempt to reduce toxicity of these compounds while retaining their bio- logical activity [4]. We present in this report informa- tion on retinoid acid esterase activity in lung and other tissues of mice.

MATERIALS AND METHODS

Methyl retinoate, ethyl retinoate, and the trimethyl- methoxyphenyl (TMMP)t analog of ethyl retinoate [ethyl all-trans-9-(4-methoxy-2,3,6-trimethylphenyl)- 3,7-dimethyl-2,4,6,8-nonatetraenoate] were prepared by Hoffmann-La Roche, Nutley, New Jersey, and Basel, Switzerland, and were supplied to us by the Lung Cancer Segment of the National Cancer Insti- tute. Diethyl p-nitrophenyl phosphate, glucose-6-phos- phate, p-nitrophenyl-N-acetyl-/J-D-glucosaminide, and p-nitrophenyl-/J-D-glucuronide were purchased from Sigma Chemical Company, St. Louis, Missouri.

For retinoid acid esterase, the standard assay mix- ture contained 0.1 ml of buffer of desired pH, 0.025 ml of 20",, ethanol containing 0.025 #mole of substrate, and enzyme preparation in a total of 0.5 ml. All buffers, potassium phosphate, pH 6.6-pH 7.4: sodium barbital, pH 8.4: sodium glycinatc, pH 9.2-pH 10.4, had an ionic strength of 0.2. Appropriate blanks lack- ing substrate or enzyme were included in the assays, all of which were performed in duplicate or triplicate.

* This study was supported by Contract N01-CP-22064 from the Division of Cancer Cause and Prevention, National Cancer Institute.

t Abbreviations used are: TMMP. trimethyl methoxy- phenyl analog.

No detectable nonenzymatic hydrolysis occurred under the assay conditions. Preparations were incu- bated for 2 hr at 37 °, and unhydrolyzed substrate was extracted by adding 2 ml of ethanol-petroleum ether (1:2, v/v). After centrifugation, the upper layer was removed and the extraction repeated with 1 ml of the same solvent. The extracts were combined and the unhydrolyzed ester measured spectrophotometrically at 370 nm. At this wavelength, the absorption due to interfering substances was minimal. In control experi- ments, the recovery was always greater than 98 per cent. Under these conditions, the solvent did not extract retinoid acids. N-Acetyl-/~-D-hexosaminidase (EC 3.3.1.30), /%glucuronidase (EC 3.2.1.31) and glu- cose-6-phosphate (EC 3.1.3.9) activities were deter- mined as previously described[5]; protein was measured by the Lowry method [6].

RESULTS AND DISCUSSIONS

Table 1 shows the distribution of esterase activity for methyl retinoate, ethyl retinoate and the TMMP analog of ethyl retinoate in mouse tissues. The data show that (a) for all three substances, brain, ovary and liver have high activity; (b) the TMMP analog of ethyl retinoate is hydrolyzed at faster rates than methyl retinoate and ethyl retinoate by brain, lung, spleen, and ovary enzymes; and (c) there is no large difference in the hydrolysis of methyl retinoate and ethyl retinoate by all tissues examined.

The results of esterase assays following differential centrifugation of homogenized lung tissue are shown in Fig. I. The subcellular distribution pattern of methyl retinoate esterase, ethyl retinoate esterase and TMMP analog of ethyl retinoate esterase are similar to that of glucose-6-phosphatase, a microsomal marker enzyme, and distinct from /3-glucuronidase and N-acetyl-/J-D-hexosaminidase, which are lysoso- mal markers. Both the retinoid acid ester hydro- lyase(s) and glucose-6-phosphatase have highest ac- tivity in the microsomal fraction; the lysosomal enzymes, /3-glucuronidase and N-acetyl-//-D-hexosa- minidase, have the highest activity in the mitochon- dria + lysosome fraction. The relative specific activity of the hydrolases in the soluble cytoplasmic fraction

947

948 C.-C. WANG and D. L. HILL

Methyl re t inoo,e esCerose

P

Ethyl ret inoo,e es~erose

> P

G - - + S (x m

B - g l u c u r o n i d o s e o ¢r

M+L L IOO

%

TMMP onolog of efhyl refinoo,e es,erose

P

lucose - 6 - 3hospho'ro se

P

M+L

D+N

- S

N-ocet~ I - / ~ - D

hexosomin idose

M+L

0 I00 pro'~eins

Table 1. Distribution of esterases for methyl retinoate, ethyl retinoate and the TMMP analog of ethyl retinoate

in mouse tissues

Tissue

Percent of substrate hydrolyzed*

TMMP analog of

Methyl Ethyl ethyl retinoate retinoate retinoate

Brain 15.6 ± 1.4 14.6 ± 0.8 18.0 ± 0.3 Ovary 13.9 ± 0.4 11.2 ± 1.0 17.5 ± 0.2 Liver 12.0 ± 1.6 14.4 _+ 3.0 11.9 ± 1.0 Kidney 7.0 ± 2.8 8.4 ± 1.9 10.0 ± 1.5 Spleen 6.0 ± 1.2 4.3 ± 1.5 10.3 ± 2.7 Lung 5.2 ± 1.1 6.8 ± 2.7 12.1 ± 0.3 Heart 3.7 ± 1.6 4.7 ± 0.8 5.7 ± 1.1 Plasma 2.5 ± 1.9 1.7 ± 0.3 1.4 ± 1.0

* DBA/2C~ mice were killed by COz suffocation and tis- sues were homogenized in 10vol. of 0.25 M sucrose. Appropriate dilutions were made so that 20 #1 contained 60 #g of protein, the amount used for assay. With larger amounts the enzymatic activity did not increase propor- tionally for all tissues. Incubations were performed as de- scribed in the text. The buffer used was phosphate, pH 7.4. Actual absorbance values for controls (incubation of substrates without enzyme) were 0.47 + 0.02 (mean ± standard error) for methyl retinoate, 0.5l ± 0.02 for ethyl retinoate and 0.51 ± 0.02 for TMMP analog of ethyl retinoate.

Fig. 1. Intracellular distribution of retinoid acid esterase and 3 marker enzymes in mouse lung. The tissues were homogenized in 9 vol. of 0.25 M sucrose and centrifuged at 2000 g for 10 min to give a nuclear fraction containing cellular debris (D + N). The resulting supernatant was cen- trifuged at 13000g for 10min to give a crude mitochon- dria-lysosome (M + L) fraction. To prepare the microso- mal fraction (P), the post M + L supernatant was centri- fuged at 105000 g for 45 rain. The supernatant constituted ioc the soluble cytoplasmic fraction. Each pellet isolated was appropriately diluted in 0.25 M sucrose so that 0.15ml contained 0.5 mg of protein, the amount used for assay, sc Similar subcellular distribution patterns were observed when lower protein concentrations (i.e., 0.3 and 0.15 rag) were used. Other conditions were as described in the text. ~ oz The buffer used was Na glycinate, pH 9.5, except for assay ; of TMMP analog of ethyl retinoate esterase, which was sodium barbital, pH 8.4. Other enzymes were assayed as o described in the text. The results are presented by the ,~ 4¢ method of DeDuve et al. [13]. Relative specific activity = percentage of total activity/percentage of total protein. On the abscissa, the fractions are represented by their relative

2C protein content. The recovery for the enzyme activities are: methyl retinoate esterase, 87.7%; ethyl retinoate esterase, 75.2%; TMMP analog of ethyl retinoate esterase, 80.6%; glucose-6-phosphate, 123°o; N-acetyl-fl-b-hexosaminidase,

91%; fl-glucuronidase, 152%; protein, 106%.

is higher than that of glucose-6-phosphate, a result suggesting the presence of an extra-microsomal ester- ase in this fraction.

The dependence of hydrolysis of retinoid esters on pH is shown in Fig. 2. Although the activities of methyl retinoate esterase in the microsomal and supernatant fractions show similar pH optima at pH 9.5, the soluble esterase activity for the T M M P ana- log of ethyl retinoate and the microsomal esterase

activity for this substrate show distinct pH-activity profiles. The supernatant enzyme has an activity opti- mum at pH 8.4; but the microsomal enzyme displays a broad activity optimum at pH 9.5-10.0.

Microsomal esterase activity for the T M M P analog of ethyl retinoate was completely inhibited by 1 pM diethyl p-nitrophenyl phosphate (Table 2). Similar

Microsomol -- , ~ i ~ "l (TMMP/RAEE)

/ . , ~ / d ' / N~So lub le (RAME) / /" \ / / "Microsomol

J / ~" (RAME}

(TMMP/RAEE)

6D 70 8.0 9.0 I00

p H

Fig. 2. Effect of pH on the hydrolysis of methyl retinoate by lung microsomal (O) and supernatant (O) esterases, and of TMMP analog of ethyl retinoate by lung microsomal (11) and supernatant (D) esterases. Tissues were homogen- ized in 3 vol. of 0.25 M sucrose, and microsomal and super- natant fractions were prepared by centrifugation as de- scribed in the legend to Fig. 1. Portions of 150pl were used for assay. Other conditions were the same as those described in the text. Abbreviations used are: RAME, methyl retinoate; TMMP/RAEE, TMMP analog of ethyl

retinoate.

Retinoid acid esterase 949

Table 2. Effect of Diethyl p-nitrophenyl phosphate on microsomal and supernatant esterase activities*

Concentration Subcellular of inhibitor % Inhibition

Substrate fraction {M) of enzyme activity

TMMP analog of ethyl retinoate

Methyl retinoate

Microsomal 1 x 10 -v 62 1 x 10 6 1130 1 × 10 -5 99

Soluble 1 x 10 -v 0 cytoplasmic 1 x 10 6 0

1 x 10 - 5 0 Microsomal 1 x 10 ~ 59

1 x 10 -6 100 1 X 10 5 100

Soluble 1 x 10 -7 57 cytoplasmic 1 × 10 -6 100

1 × 10 5 95

* Enzyme preparation and assay conditions were the same as for Fig. 3, except the enzymes were preincubated with buffer and inhibitor for 30min at 0 ° before the addition of substrates.

results were obtained with bo th microsomal and supernatant methyl ret inoate esterases. In contrast, the soluble esterase activity for the T M M P analog of ethyl ret inoate was insensitive to the inhibi tor even at a concentra t ion of 10/~M. The microsomal and supernatant esterases hydrolyzing the T M M P analog of ethyl ret inoate can also be differentiated by their thermal stability, as revealed by heating the enzymes at 52 ° (Fig. 3A). The microsomal enzyme loses about 25 per cent of its activity after 10 min of incubat ion but retains the remaining activity for up to 2 hr. In contrast, activity of the supernatant esterase initially increases about 10 per cent, but this is followed by a decrease to about 50 per cent of its original activity.

The difference in pH optima, sensitivity to inhibi- tor, and thermal stability demonst ra te the presence in mouse lungs of different microsomal and superna- rant forms of enzymes hydrolyzing the T M M P analog of ethyl retinoate. As a further indication of a possible multiplicity of esterases acting upon retinoid esters, there is also a difference in the thermal stability of microsomal and supernatant methyl ret inoate ester-

ases (Fig. 3B). The microsomal enzyme activity has a stability similar to that of microsomal enzyme hyd- rolyzing the T M M P analog of ethyl retinoate, losing about 30 per cent of its activity after 10 min of incu- bat ion but retaining the remaining activity for up to 2 hr. In contrast, activity of the supernatant methyl ret inoate esterase initially increases about 10 per cent, stays at this level for 20 min, then gradually de- creases to about 70 per cent of original activity after 2 hr of incubation.

These data indicate the presence in mouse lung of at least one microsomal and two soluble esterases uti- lizing retinoid esters as substrates. The main microso- mal localization of ret inoid acid esterase is similar to those of retinyl acetate hydrolase [7] and non-spe- cific esterase [8] but is distinct from the nucleic and mitochondrial- lysosomal localization of retinyl palmi- tate hydrolase [7].

By the use of inhibitors, enzyme fractionation and a compar ison of enzyme properties, Ganguly [-9] con- cluded that liver retinyl acetate esterase, non-specific esterase, and cholesterol esterase are independent

6 E

o

120[ ~p~ Subst"ro~'e ' [1~ \ TMMP analog of

IO0~ X efhyl refinoate

80 ~ . . . . .

6O

4 0 ~ 2 0 Soluble

I (°)1 I I I I I 0 20 40 60 80 tOO 120 0

Time, min

~ Substrol-e:

ic M e f h Y r ~

(b) I I I I I 20 40 60 80 IO0

Time, min

120

Fig. 3. (A). Heat inactivation of lung microsomal (0) and soluble (O) esterase for TMMP analog of ethyl retinoate at 52 °. The enzyme fractions were prepared as described in the legend to Fig. 2. Following the heat treatment, the mixtures were cooled in ice and analyzed for esterase activity. The assay conditions were as described in the legend to Fig. 2, except that sodium barbital, pH 8.4, was used as the buffer. (B). Heat inactivation of lung microsomal (0) and soluble (O) esterase for methyl retinoate at 52 °. The enzymes were assayed at pH 9.5. Other details were as for Fig. 3A.

950 C.-C. WANG and D. L. HILL

entities. On the other hand, it was recently shown that retinyl acetate is also a substrate for a highly purified non-specific esterase from liver and kid- ney [10]. It is questionable, therefore, whether the dis- t inct ion between retinyl acetate esterase and non- specific esterase is still justified. We do not know from the present work whether ret inoid acid ester hydro- lase and non-specific esterase are similar or indepen- dent enzymes. However, since the non-specific ester- ase is known to exist in multiple forms, it is likely tha t one or more of its forms participates in the hy- drolysis of ret inoid acid esters.

The occurrence of an unidentified, "physiological" ester of retinoic acid in all tissues of rats [11, 12] and the present finding that ret inoid acid esterase activity is present in all tissues suggest that the enzymes have a role in the action of retinoic acid.

REFERENCES

1. M. V. Cone and P. Nettesheim, d. natn. Cancer Inst. 50, 1599 (1973).

2. U. Saffiotti, R. Montesano, A. R. Sellakumar and S. A. Borg, Cancer 20, 857 (1967).

3. M. B. Sporn, N. M. Dunlop, D. L. Newton and J. M. Smith, Fedn Proc. 35, 1332 (1976).

4. W. Bollag, Eur. J. Cancer 10, 731 (1974). 5. C. C. Wang, S. Straight and D. L. Hill, Biochem. Phar-

mac. 25, 471 (1976). 6. O. H. Lowry, N. J. Rosebrough, A. L. Farr and R.

J. Randall, J. biol. Chem. 193, 265 (1951}. 7. S. Mahadevan, N. I. Ayyoub and O. A. Roels, J. biol.

Chem. 241, 57 (1966). 8. A. Omachi, C. P. Barnum and D. Glick, Proc. Soc.

exp. Biol. Med. 67, 133 (1948). 9. J. Ganguly, J. Sci. Ind. Res. 26, 110 (1967).

10. J. Bertram and K. Krisch, Eur. J. Biochem. II, 122 (1969).

11. Y. L. Ito, M. Zile, H. F. DeLuca and H. M. Ahrens, Biochim. biophys. Acta 369, 338 (1974).

12. Y. L. Ito~ M. Zile, H. Ahrens and H. F. DeLuca, J. Lip. Res. 15, 517 (1974).

13. C. DeDuve, B. C. Pressman, R. Gianetto, R. Wattiaux and F. Appelmans, Biochem. J. 60, 604 (1955).